search

MarineRoboticsGroup / dcsam

Factored inference for discrete-continuous smoothing and mapping.
https://arxiv.org/abs/2204.11936
MIT License
143 stars 18 forks source link

DCMixtureFactor normalization #33

Closed plusk01 closed 1 year ago

plusk01 commented 1 year ago

The normalization was returning +log(eta) instead of -log(eta), where eta is the normalization constant for a multivariate Gaussian.

Deleting the DCMixtureFactor.h specialization of nonlinearFactorLogNormalizingConstant allowed the calculation to fall back to the correctly implemented function in DCFactor.h.

I added a new test in dccontinuous_mixture that now passes.

keevindoherty commented 1 year ago

Great catch, thanks @plusk01

The change looks good to me :)